Comment by aumfer on Who Wants To Start An Important Startup? · 2012-09-07T18:52:18.877Z · LW · GW

That religions tend to confuse a factual issue (the existence or non-existence of various divine superbeings and their various characteristics) with moral issues, is one of the problems that I'd like to see solved, not contribute to its confusion.

That's precisely what I was hoping to do. Analysis could show the relationship between users' identification as atheist/religious and their surveyed moral attitudes. Presumably this relationship might not be nearly as strong as people think.

Comment by aumfer on Who Wants To Start An Important Startup? · 2012-09-07T13:21:36.733Z · LW · GW

So you're saying that irrational beliefs aren't harmful?

I thought the point of this site was to help people develop a system to reduce exactly such non-predicting "beliefs."

One of the things that I feel discourages non-atheists from even questioning their belief is that they see atheists as "non-moral." That is, morality doesn't enter into the equation of atheism (it's only belief/non-belief), but it is intimately tied to religion. Codifying the moral beliefs of atheists and believers could help to promote atheism as an alternative to religion. Not to mention the interesting data that could be obtained about moral values as related to self-described religious identity.

Comment by aumfer on Who Wants To Start An Important Startup? · 2012-09-06T19:17:23.643Z · LW · GW

Abolish Atheism

The way to reduce the harmful power of religion isn't by getting people to stop believing in God. It's by getting them to codify and share whatever their moral belief system is. Atheism doesn't work because it doesn't really say anything. How often do you hear the phrase "most atheists" when discussing the morality of atheism?

Put together (with the help of some theologists/philosophers/social scientists/etc) a short but comprehensive quiz covering a variety of moral topics.

Once a good amount of data has been collected, apply some unsupervised machine learning classification to the dataset to determine "moral groupings" based on the algorithmically-determined relationships.

From here it would be interesting to

1. Compare these groups to the self-identified religions of the participants
2. Use these groups as alternatives to traditional religions/atheism as they provide a picture of moral leanings that is less biased (based purely on data rather than "history" or "branding).

Ok so that ignores the profitable criteria, but it's in my "google doc o' ideas", and I'd love to be able to analyze a dataset like that.

Note: could also apply a similar idea to political parties